High Voltage Power Supply Units

Popular Tesla Coil PSUs

Coilers use a wide variety of HV PSUs.  Here is a quick summary from smallest to largest.
Currently, this page will only go into detail on NSTs.  More PSUs will be covered here as I gain experience.

OBIT - oil burner ignition transformer, puny
MOT - microwave oven transformer, typically voltage is a little low and you have to make series/parallel connections
NST - neon sign transformer, ideal for coils with up to 5 ft arc output, easily damaged by TC service
PT - potential transformer, requires ballast circuitry 
PIG - distribution transformers like the ones you see on light poles.  Most coilers use  5-20KVA, 14KV units.  Requires ballast circuitry

Neon Sign Transformers (NST)

This is what you might find at your local Sign Shop I just got my hands on these 60 ma NSTs.  I found them at a small shop that makes Neon Signs and does sign repair in general.  From many months of searching I found that small shops are the best and DON'T rely on phone calls.  There are basically 4 types of NSTs - you have 12KV models and 15KV models.  Each voltage comes in either 30ma or 60ma.  Look for the 60ma models, but don't pass up the 30's.  Richard Quick has an excellent article on all aspects of NST usage.   I will look for it and link it here. Richard Quick Article on NSTs This article covers depotting!  I'm an order of magnitude more agressive than Richard when It comes to increasing current output, but he is trying to build a reliable device and I'm just playing around on the upper limits of performance in the name of experimentation.  Remember, many NSTs die at their normal operating current (much less my 200+ ma!).

Enhancing an NST

NSTs are current limited transformers.  This means that a NST will only produce a fixed amount of current regardless of the output load.  The current is limited by using packs of metal shunts to "short circuit" the flux from the primary and prevent it from intersecting the secondary.  The current limit may be adjusted by adding or removing individual leaves in these shunts.  The following pics are probably much more helpful than my ramblings...

All pics are "clickable"

	Final version Final version of the NST enhancement.  Note the acrylic "C" shaped parts that completely wrap around the secondaries.  I also coated the edges of the secondaries with about 2 coats of clear epoxy.  Testing shows that these additions, with the additional precaution of reducing my input voltage to 120V, eliminates all of the nasty flashover and corona problems that I was having.
	Here is a Franceformer 12KV, 60ma unit after LOTS of work.  The primary is the heavy gauge coil in the center.  The secondaries are wound on each side.  You can see the HV end of the secondaries going off to the banana binding posts on the plexi and you can see the ground end of the secondaries running out to a bolt on the E core.   The shunts are located on each side of the primary and contain about 8 leaves each in this photo.  I have the shunts held in with some scraps of LDPE for testing and photos.  I need to investigate a more secure way to hold them in place for actual TC use. Note the insulative tape wound on each side of the core nearest the secondary.  I need to revise the insulation to include the top and bottom of the core as well, occasionally an arc will creep along the surface of the tape to the core! Note this configuration is for photo/educational purposes only.  Flashover and corona were severe when operated in this configuration.  See "final version" photo above
	This is a top view showing the primary connection, 2 HV connections, and the ground connection in the center.  Note that the core is composed of (2) E shaped sections that are held together by 4 bolts on the corners.  There is no weld in the seam so you can easily seperate the halves and remove the primary and secondaries. Note this configuration is for photo/educational purposes only.  Flashover and corona were severe when operated in this configuration.  See "final version" photo above
	Close-up of the shunts. Note this configuration is for photo/educational purposes only.  Flashover and corona were severe when operated in this configuration.  See "final version" photo above

Revision B (see previous work below)

Although I couldn't get it to arc over at 140V input open circuit, it displays some rather nasty flash-overs when integrated with the RC filter and spark gap.  The flash over usually occurs when trying to ignite the gap.  I coated both the inboard and outboard edges of the NST secondary coils with epoxy.  I made some acrylic "C" shaped parts that wrap around the secondaries.   The NST originally had a tar paper version of this when it was potted.

I decided to stick with lower voltages and run the NST at its rated input.  I rewired my control box to only supply 120 VAC.  I adjusted the shunts in the NST to allow for about 200ma at the lower 116V input voltage.  Between the lower voltages and better insulation, I hope that cured the flash-over problem.

Here is the new test data

Vin was measured with a Wavetek 27XT.  (I out) was measured with a Simpson 200ma analog meter.  (I in) was measured with the cheap current meter on my control box.  It's not very accurate but is handy in showing a trend.  My output current goes up to a solid 200ma and the input current drops to 20.5A with the addition of the 135uF PFC cap.

Revision A (see previous work below)

The original configuration displayed occasional flash-over problems when running the variac at settings above 120V.  Had a problem with the NST HV bushings arcing to an angle bracket used to support the plexi deck.  I removed the angle bracket.  I removed the Kapton tape from the sides of the E core and replaced it with electical tape and scraps of PE from a milk jug.

 Preliminary testing of the de-potted NST

Voltage Input was 134-136VAC from my controller

Results of adding a 135uF PFC capacitor to my ENST on the "8 leaves / shunt" setting
Note that the PFC cap almost halved the imput current while giving me slightly more output current.
I believe that the gain in output current happened b/c the input voltage rose a bit with the lower  current demand.

I wanted to check for saturation of the NST core at higher power levels.
Saturation should be evidenced by a steep increase in input current with respect to output current.
The NST has 8 leaves in each of the 4 shunts and is equipped with the 135uF PFC.

If you plot Output Current vs Input Current, you get a nice linear relationship. 
 This suggests that saturation is not a problem at this power level.

[ Back to Tesla Main   |  Back to Ross Home ]